Simultaneous extraction of geometry and surface properties of targets using simple infrared sensors

Abstract

We investigate the use of low-cost infrared (IR) sensors for the simultaneousextraction of geometry and surface properties of commonly encounteredfeatures or targets in indoor environments, such as planes, corners,and edges. The intensity measurements obtained from such sensorsare highly dependent on the location, geometry, and surface propertiesof the reflecting target in a way that cannot be represented by asimple analytical relationship, therefore complicating the localizationand recognition process. We propose the use of angular intensityscans and present an algorithm to process them to determine the geometryand the surface type of the target and estimate its position. Themethod is verified experimentally with planes, 90-deg corners, and90-deg edges covered with aluminum, white cloth, and Styrofoam packagingmaterial. An average correct classification rate of 80% of both geometryand surface over all target types is achieved and targets are localizedwithin absolute range and azimuth errors of 1.5 cm and 1.1 deg, respectively.Taken separately, the geometry and surface type of targets can becorrectly classified with rates of 99 and 81%, respectively, whichshows that the geometrical properties of the targets are more distinctivethan their surface properties, and surface determination is the limitingfactor. The method demonstrated shows that simple IR sensors, whencoupled with appropriate processing, can be used to extract substantiallymore information than that for which such devices are commonly employed.(C) 2004 Society of Photo-Optical Instrumentation Engineers.